Languages listed in ISO 639: feedback sought

[Mark Mandel]

>From LINGUIST List #11-230
Please send replies not to me but to
   John Clews <Endanger at sesame.demon.co.uk>
Do not simply post them to SLLING-L; he does not read it.

I attach an extract from a request for comments on and contributions to a list
of codes for representing the names of languages, ISO [International Standards
Organization] Standard #639. It was posted on the LINGUIST List. You can find
the full text on the Web at
     http://linguistlist.org/issues/11/11-230.html

The reason I am posting it here is that the only mention of sign languages in
the list is a single item, indicating that the Library of Congress uses the
3-letter code "sgn" for "Sign languages", without further distinction between
them, and that neither of the existing versions of the ISO standard has anything
at all for sign languages. (Please do not blame or flame Mr. [Dr.?] Clews. He is
not responsible for this list!)

I have written to him as follows:

     "Sign languages (not expanded further)" is about as acceptable as
     would be "Asian languages (not expanded further)". I will start with
     proposing
          asl  American Sign Language
     and continue by posting a message about your request on SLLING-L,
     the Sign Language Linguistics List, to elicit contributions from sign
     linguists familiar with others of the dozens or hundred-plus of known
     sign languages in the world.

I strongly suggest to sign linguists who are specialists in or familiar with
other sign languages that you email your suggestions for codes for those
languages as soon as possible. The
     Joint Advisory Committee on ISO 639: Codes for representation of names of
languages
will be meeting in Washington, DC, February 17-18, and I suspect that
information should be fed to them well in advance of those dates.

I also suggest that you look at the web site listed above for the existing list
of language codes. It is possible that the usual sign linguists' abbreviation
for a particular SL is already established in use for a spoken language, and
some alternative code will have to be found for the SL. This has already been
done with many spoken languages in the list, such as this set:
  LC  ISO 639-2   ISO 639-1  Language name in English
- ----------------------------------------------------------
      ara          ar        Arabic
      arc                    Aramaic
      arp                    Arapaho
      arn                    Araucanian (Mapuche)
      arw                    Arawak

A number of entries on the list refer to sets of languages, but almost all of
these entries are for sets of *related* languages, such as
      apa                    Apache languages
While it might make sense to have a listing for "French Sign Language and
related SLs", each of those languages should also be listed; and other SLs, such
as Japanese and Hong Kong SLs, could not be included in it.

The main purpose of this code is for use in computer systems. While most sign
languages have no written form, that situation is changing rapidly with systems
like SignWriting that are intended for signers to use. There are also systems
like Stokoe notation and HamNoSys that are used by sign linguists.

Sincerely,
Mark Mandel

   Mark A. Mandel : Senior Linguist and Manager of Acoustic Data
         Mark_Mandel at dragonsys.com : Dragon Systems, Inc.
 320 Nevada St., Newton, MA 02460, USA : http://www.dragonsys.com/
                     (speaking for myself)

     >>>>>  EXTRACT FROM LINGUIST LIST ANNOUNCEMENT >>>>>

Dear list members. I am a member of the Joint Advisory Committee on
ISO 639: Codes for representation of names of languages (abbreviated
to ISO 639: language codes in further discussion below).

This committee meets on 17-18 February 2000 in Washington DC, and I
would be grateful for any information from list members, which would
highlight any major gaps in  the languages listed below.  The major
interest at this stage is in written languages.

If you are interested in the background to ISO 639, read section 1:
if not, and you would like to comment on the codes, and the languages
that have been coded, and any omissions or errors, go to section 2.

1. ISO 639: language codes

ISO 639 is one of many international standards developed by groups of
experts in many countries. This section provides a simplified view.

Put simply, ISO 639's job is to provide simple codes that can be
embedded in (mainly computerised) information systems that can allow
these information systems to highlight language use, or even to
enable useful things like font switching or similar, e.g. on Internet
web sites.

There are older 2-letter codes used, it could be said, mainly in
older, "legacy" system. 3-letter codes (mainly identical with codes
used by the Library of Congress, and in many libraries) have seen the
largest growth, and allow for greater expansion. Actually the two
sets are currently listed in two separate parts of ISO 639,
respectively in ISO [WD] 639[-1] and ISO 639-2.

The Internet Engineering Task Force's specification RFC 1766
recommends the use of ISO 639 codes in Internet uses.

     [...]

2. Opportunity for feedback

The list [see web site-- MAM] is my own handy reference list based on my own
compilation of 3-letter codes from ISO 639-2 and Library of Congress
codes, and the 2-letter codes in ISO WD 639-1. Errors are likely to
be my own, rather than in ISO 639, though I have been fairly careful.

     [...]

I'd be particularly interested to know of obvious omissions, or
errors in naming, or where predominant use of language names has
changed.

There are also some fairly basic "genetic codes" where entries for
"xxxx languages" or "xxxx languages (other)." Again, if some language
groups seem to have been omitted altogether I would be glad to know.

     [...]

I am primarily interested in omissions, and in language names, but if
you want to suggest a particularly useful 3-letter code, it may be
helpful (although I think that the default approach is likely to be
"use the first three letters of the name in English, or French, or in
the local spelling of the name" if known, allowing for transcription
or transliteration conventions).

     [...]